Leveraging Generative AI for Security Scanning

No ratings

Presented at BSidesSATX 2026 by

This talk will show how Gen AI can reason about these problems in ways that current tools can't. We'll walk through vulnerable code examples across three categories: authentication, authorization, and secure design. You'll see where Gen AI excels, fails, and how to get measurable results from a technology that is inherently inconsistent. You'll leave with concrete prompting strategies and a methodology for turning unpredictable AI output into something your team can actually act on.